International Journal of Molecular Sciences

Article Interspecies Outer Membrane Vesicles (OMVs) Modulate the Sensitivity of Pathogenic and Pathogenic Yeasts to Cationic Peptides and Serum Complement

Justyna Roszkowiak 1 , Paweł Jajor 2 , Grzegorz Guła 1, Jerzy Gubernator 3, Andrzej Zak˙ 4 , Zuzanna Drulis-Kawa 1 and Daria Augustyniak 1,*

1 Department of Pathogen Biology and Immunology, Institute of Genetics and Microbiology, University of Wroclaw, 51-148 Wroclaw, Poland; [email protected] (J.R.); [email protected] (G.G.); [email protected] (Z.D.-K.) 2 Department of Pharmacology and Toxicology, Faculty of Veterinary Medicine, Wroclaw University of Environmental and Life Sciences, 50-375 Wroclaw, Poland; [email protected] 3 Department of Lipids and Liposomes, Faculty of Biotechnology, University of Wroclaw, 50-383 Wroclaw, Poland; [email protected] 4 Ludwik Hirszfeld Institute of Immunology and Experimental Therapy, Polish Academy of Science, 53-114 Wroclaw, Poland; [email protected] * Correspondence: [email protected]; Tel.: +48-71-375-6296

 Received: 30 September 2019; Accepted: 5 November 2019; Published: 8 November 2019 

Abstract: The of bacterial outer membrane vesicles (OMVs) contributes to innate microbial defense. Limited data report their role in interspecies reactions. There are no data about the relevance of OMVs in bacterial-yeast communication. We hypothesized that model Moraxella catarrhalis OMVs may orchestrate the susceptibility of pathogenic bacteria and yeasts to cationic peptides () and serum complement. Using growth kinetic curve and time-kill assay we found that OMVs protect Candida albicans against polymyxin B-dependent fungicidal action in combination with fluconazole. We showed that OMVs preserve the virulent filamentous phenotype of yeasts in the presence of both antifungal drugs. We demonstrated that bacteria including Haemophilus influenza, Acinetobacter baumannii, and Pseudomonas aeruginosa coincubated with OMVs are protected against membrane targeting agents. The high susceptibility of OMV-associated bacteria to polymyxin B excluded the direct way of protection, suggesting rather the fusion mechanisms. High-performance liquid chromatography- spectroscopy (HPLC-UV) and zeta-potential measurement revealed a high sequestration capacity (up to 95%) of OMVs against model cationic peptide accompanied by an increase in surface electrical charge. We presented the first experimental evidence that bacterial OMVs by sequestering of cationic peptides may protect pathogenic yeast against combined action of antifungal drugs. Our findings identify OMVs as important inter-kingdom players.

Keywords: outer membrane vesicles (OMVs); Candida albicans; antimicrobial peptides; complement; interspecies interactions; inter-kingdom protection; fungicidal activity; fluconazole; hyphae

1. Introduction In the environment, different microbial populations, including bacteria and fungi, co-exist. Mixed microbial populations are also a common feature in many diseases. Hence, understanding which factors may be potentially important players in interspecies dynamics of growth and to what degree these dynamics are mediated by the host is very important. Among these factors are outer

Int. J. Mol. Sci. 2019, 20, 5577; doi:10.3390/ijms20225577 www.mdpi.com/journal/ijms Int. J. Mol. Sci. 2019, 20, 5577 2 of 21 membrane vesicles (OMVs) of Gram-negative bacteria classified recently as secretion system type zero [1]. These proteoliposomal nanoparticles released from the cell play an important role in bacterial physiology and pathogenesis. During pathogenesis, they enhance cellular adherence, cause biofilm formation, and induce apoptosis or inflammation [2–4]. Furthermore, DNA-containing OMVs, through horizontal gene transfer, are important in interspecies communication as well as in host–pathogen interactions [5]. OMVs contribute also to innate bacterial defense through β-lactamase content or trapping and degrading the membrane-active peptides [6–10]. Accordingly, the involvement of OMVs in resistance to with various modes of action is growing [11]. Both offensive and defensive functions of OMVs have been reported. In line with this, OMVs can deliver bactericidal toxins or enzymes to other bacteria. Lytic activities have been documented for OMVs derived from Pseudomonas aeruginosa or Myxococcus xanthus [12–14]. On the other hand, for example, M. catarrhalis OMVs carrying β-lactamases confer protection to M. catarrhalis and Streptococcus pneumoniae against β-lactam antibiotics [15]. Likewise, OMVs from β-lactam-resistant E.coli can protect β-lactam-susceptible E.coli and fully rescue them from β-lactam -induced growth inhibition [16]. Several earlier studies have shown that OMVs are involved in the trapping of antimicrobials, thus providing protection among bacteria, essentially in intraspecies systems [6,8,17]. Nevertheless, limited data are available on the relevance of this phenomenon in various interspecies populations. Among mixed populations, the degree of OMV-dependent protection of individual partners can be completely different. One of the factors that may influence this is the physicochemical nature of the vesicle itself. It includes both the intrinsic antimicrobial binding capacity, which is the result of inherited or acquired resistance [18] but also the ability of vesicle, based on physicochemical properties, to interact with other target cells. There are several factors responsible for the latter phenomenon including electric charge and hydrophobicity of a cellular surface [19,20]. The metabolic activity of OMV-producing bacteria also seems to have a significant impact on this issue. For example, for P. aeruginosa it was reported that variation in physicochemical properties was dependent on growth phase (exponential versus stationary), influencing, therefore, the cell association activity [19]. All aforementioned characteristics of OMVs indicate the significant degree of selectivity in OMV–cell interactions. Therefore, considering variety of microbial populations, the question of which bacteria are and which are not protected against membrane-active agents is still very ambiguous. It is also of great importance to ask whether this protection can go beyond the protection against only bacteria, affecting pathogens from other kingdoms such as fungi. Hence, it is of interest to investigate the protective potency of OMVs in the light of their specificity to react with a target cell. In the present study, we used our well-characterized OMVs from Moraxella catarrhalis Mc6 [4,21] as model vesicles and characterized their protective potential against model membrane-active agents in various interspecies combinations. First, we examined the protective activity of OMVs in bacterial intra- and interspecies systems and mode of observed protection, showing by HPLC-UV, zeta potential measurement, and o-nitrophenyl-β-D-galactopyranoside ONPG-based permeabilization assay, highly effective sequestration. Next, we examined the protective potential of OMVs in the bacterial-yeast inter-kingdom system. As far as we know, our study is the first report that OMVs can protect Candida albicans against antifungals drugs. Finally, we investigated whether OMVs-dependent protection is a signature of only free vesicles or those associated with the cell.

2. Results

2.1. Mc6 OMVs Characteristics As shown in transmission electron microscopy (TEM) image, the diameters of outer membrane vesicles from M. catarrhalis 6 had 30–200 nm (Figure1a). The results of measurements of OMV particle size/zeta potential are shown in Figure1b. The protein and lipooligosaccharide (LOS) components of OMVs are shown in Figure1c,d, respectively. As we documented previously [ 21], the pivotal outer membrane proteins packaged in these vesicles were OmpCD, OmpE, UspA1 (ubiquitous surface protein A1, Hag/MID (Moraxella immunoglobulin D-binding protein), CopB, MhuA Int. J. Mol. Sci. 2019, 20, 5577 3 of 21 Int. J. Mol. Sci. 2019, 20, x FOR PEER REVIEW 3 of 20 binding(hemoglobin-binding protein) , TbpA protein), (transferrin-binding TbpA (transferrin-binding protein A), proteinTbpB (transferrin-binding A), TbpB (transferrin-binding protein B), protein LbpB (lactoferrin-bindingB), LbpB (lactoferrin-binding protein B), protein OMP M35, B), OMP and M35,MipA and (structural MipA (structural protein). protein).

FigureFigure 1. 1. PhysicalPhysical characterization characterization of ofouter outer membrane membrane vesicles vesicles (OMVs) (OMVs) released released from from Mc6 Mc6 cells: cells: (a) (a) TEM image of OMVs, vesicles are indicated by arrows (magnification, 50,000); (b) the size TEM image of OMVs, vesicles are indicated by arrows (magnification, ×50,000);× (b) the size distributiondistribution byby volumevolume and and the the zeta zeta potential potential of vesicles, of vesicles, as assessed as assessed by the Zeta-sizer; by the Zeta-sizer; each experiment each experimentwas performed was inperformed triplicate; in (c) triplicate; the respresentative (c) the respresentative proteinogram proteinogram of 12% SDS-PAGE of 12% electrophoresis SDS-PAGE electrophoresisof OMVs; the protein of OMVs; profiles the wereprotein visualized profiles usingwere Coomassievisualized staining;using Coomassie (d) the representative staining; (d) 10%the representativeSDS-PAGE electrophoresis 10% SDS-PAGE of lipooligosaccharide electrophoresis of (LOS)-OMVs lipooligosaccharide visualized (LOS)-OMVs by silver staining. visualized by 2.2. M.silver catarrhalis staining.OMVs Passively Protect Cross-Pathogens against Polymyxin B-Dependent Killing

2.2. M.Polymyxin Catarrhalis B OMVs (PB) was Passively used as Protect a model Cros ofs-Pathogens cationic peptide. against In Polymyxin our study, B-Dependent using 4-h time-kill Killing assay, the minimal bactericidal concentrations (MBC) of PB against 5 105–106 cfu/mL of prominent human Polymyxin B (PB) was used as a model of cationic peptide.× In our study, using 4-h time-kill pathogens, including nontypeable Haemophilus influenzae, Pseudomonas aeruginosa, and Acinetobacter assay, the minimal bactericidal concentrations (MBC) of PB against 5 × 105–106 cfu/mL of prominent baumannii, was in the range 0.5–5 µg/mL, and caused killing effect within 2 h (Figure2a). When human pathogens, including nontypeable Haemophilus influenzae, Pseudomonas aeruginosa, and pathogenic bacteria were incubated with a bactericidal concentration of PB in combination with Acinetobacter baumannii, was in the range 0.5–5 μg/mL, and caused killing effect within 2 h (Figure 20 µg/mL OMVs from Mc6, the bacteria showed active growth in contrast to the antibiotic alone, which 2a). When pathogenic bacteria were incubated with a bactericidal concentration of PB in combination remained at the level of control. Up to 10 times lower concentrations of OMVs had no effect or the with 20 μg/mL OMVs from Mc6, the bacteria showed active growth in contrast to the antibiotic alone, effect was negligible. The amount of OMVs that was required to achieve complete protection against which remained at the level of control. Up to 10 times lower concentrations of OMVs had no effect or respiratory pathogens referred to 20 µg/mL. The higher concentrations of vesicles did not intensify the effect was negligible. The amount of OMVs that was required to achieve complete protection the growth. against respiratory pathogens referred to 20 μg/mL. The higher concentrations of vesicles did not intensify2.3. OMVs the Protect growth. Serum-Sensitive Strains and Accelerate Growth of Serum-Resistant Strains against Complement 2.3. OMVs Protect Serum-Sensitive Strains and Accelerate Growth of Serum-Resistant Strains against To define the influence of OMVs on the survival of cross-pathogens in the presence of active Complement serum (NHS), the 4 h complement bactericidal assays were performed (Figure2b). Of the three studiedTo define cross-pathogens, the influence only of nontypeable OMVs on theH. survival influenzae ofappeared cross-pathogens to be serum-sensitive. in the presence Two of others,active serumA. baumannii (NHS),and theP. 4 aeruginosa, h complementwere serum-resistant. bactericidal assays The resistantwere performed strains showed (Figure either 2b). 100%Of the survival three studiedor only across-pathogens, slight decrease (withinonly nontypeable one order) H. after influenzae 4 h of incubation appeared to in be 50% serum-sensitive. or 75% NHS. Nontypeable Two others, A.H. influenzaebaumanniistrains and P. were aeruginosa, sensitive were to NHS-dependent serum-resistant. killing The withresistant a reduction strains ofshowed cfu/mL either in the range100% survivalof six orders or only after a 30slight min decrease of incubation (within (NTHi3) one or order) 120 after min of4 h incubation of incubati (NTHi6)on in 50% in the or presence75% NHS. of Nontypeable25% or 50% serum, H. influenzae respectively, strains compared were sensitive to the initialto NHS-depen inoculum.dent These killing strains with co-incubated a reduction with of cfu/mLNHS and in the 20 µ rangeg/mL of of six OMVs orders exhibited after 30 amin time-dependent of incubation (NTHi3) increase or in 120 survival, min of which incubation was similar (NTHi6) to ingrowth the presence of control of incubated25% or 50% in theserum, presence respective of heat-inactivatedly, compared to serum the initial (HiNHS). inoculum. These strains co-incubated with NHS and 20 μg/mL of OMVs exhibited a time-dependent increase in survival, which was similar to growth of control incubated in the presence of heat-inactivated serum (HiNHS).

Int. J. Mol. Sci. 2019, 20, 5577 4 of 21 Int. J. Mol. Sci. 2019, 20, x FOR PEER REVIEW 4 of 20

FigureFigure 2. 2.Mc6 Mc6 OMVs OMVs protect protect bacteria bacteria fromfrom otherother speciesspecies against against polymxin polymxin B-dependent B-dependent and and human human complement-dependentcomplement-dependent bactericidal bactericidal activity:activity: ( (aa)) Polymyxin Polymyxin B-dependent B-dependent killing; killing; (b ()b human) human serum serum (NHS)(NHS) complement-dependent complement-dependent killing. killing. BacteriaBacteria from log log phase phase were were incubated incubated for for 4 h 4 in h inthe the presence presence ofof indicated indicated concentrations concentrations of of membrane-targeting membrane-targeting ag agentsents alone alone or or along along with with OMVs OMVs and and plated plated in 0, in 0, 30,30, 120, 120, and and 240 240 min. min. Data Data are are expressed expressed asas meanmean cfucfu/mL/mL ± SDSD from from three three independent independent experiments experiments ± performedperformed in in triplicate. triplicate.

A.A. baumannii baumanniiincubated incubated onlyonly inin thethe presencepresence of of active active NHS NHS was was slightly slightly killed killed by by the the lytic lytic complementcomplement action action within within 4 4 h, h, whereas whereas after after parallel parallel exposure exposure toto OMVs,OMVs, itsits growthgrowth waswas at the level of HiNHSof HiNHS control. control.P. aeruginosa P. aeruginosawas was highly highly resistant resistant to to75% 75% actionaction of serum, serum, thus thus co-incubation co-incubation with with OMVsOMVs did did not not change change the the growth growth dynamicsdynamics (data not shown). shown). When When control control bacteria bacteria were were incubated incubated inin the the presence presence of of heat-inactivated heat-inactivated NHS,NHS, neitherneither killing nor nor other other growth growth alterations alterations were were observed observed

Int.Int. J. Mol.J. Mol. Sci. Sci.2019 2019, 20, 20, 5577, x FOR PEER REVIEW 5 of5 20 of 21

for any of the tested strains, indicating that observed lysis was complement-dependent. Overall, these forfindings any of theindicate tested that strains, OMVs indicating efficiently thatprotect observed serum-sensitive lysis was gram-negative complement-dependent. pathogens against Overall, thesecomplement findings indicateaction while that accelerating OMVs efficiently the gr protectowth of serum-sensitivemoderately serum-resistant gram-negative strains. pathogens against complement action while accelerating the growth of moderately serum-resistant strains. 2.4. Mc 6 OMVs Passively Protect Pathogenic Yeasts against Polymyxin B-Dependent Fungicidal Effect in 2.4.Combination Mc 6 OMVs with Passively Fluconazole Protect Pathogenic Yeasts against Polymyxin B-Dependent Fungicidal Effect in Combination with Fluconazole It has been previously reported that the antifungal effect of polymyxin B combined with fluconazoleIt has been can previously be synergistic reported or potentiated that the antifungal [22]. We e ffthereforeect of polymyxin initially examined B combined the with sensitivity fluconazole of canCandida be synergistic albicans orto potentiatedcombined action [22]. of We both therefore drugs. initially Our results examined showed, the that sensitivity for ∼10 of5 cfu/mLCandida of albicans this 5 toyeast, combined the MIC action50 for of fluconazole both drugs. (FLC) Our and results MIC showed,100 for polymyxin that for ~10B (PB)cfu were,/mL respectively, of this yeast, 1 the μg/mL MIC 50 forand fluconazole 128 μg/mL. (FLC) Next, and based MIC on100 checkerboardfor polymyxin assay, B (PB) we assessed were, respectively, the growth of 1 µC.g /albicansmL and incubated 128 µg/mL. Next,with based polymyxin on checkerboard B alone, fluconazole assay, we assessedalone, or the their growth combinations of C. albicans at variousincubated concentrations. with polymyxin The B alone,results fluconazole revealed that alone, polymyxin or their B combinations at concentratio atns various much lower concentrations. than MIC (1/8 The MIC results and revealed 1/16 MIC) that polymyxinexerts a potent B at concentrations antifungal effect much against lower C. than albicans MIC (1when/8 MIC combined and 1/16 with MIC) 1 exertsμg/mL a (MIC potent50) antifungalof FLC eff(Figureect against 3a). C. albicans when combined with 1 µg/mL (MIC50) of FLC (Figure3a).

FigureFigure 3. 3.OMVs OMVs protect protect yeast yeast againstagainst synergisticsynergistic fungicidal activity activity of of polymyxin polymyxin B Bin incombination combination withwith fluconazole: fluconazole: ((aa)) SynergisticSynergistic activities activities determin determineded by bycheckerboard checkerboard assay; assay; (b) time-kill (b) time-kill assay; assay;(c) (c)24 24 h h growth-curve growth-curve kinetics kinetics for for C.C. albicans albicans incubatedincubated with with drug drug alone alone and and drug drug in incombinations combinations with with oror without without OMVs. OMVs The. The kinetics kinetics werewere measured using using Varioskan™ Varioskan™ LUXLUX read readerer with with measurements measurements at at 1 h1 intervals.h intervals. The The data data for for experiments experiments (a,b)(a,b) show means ± SDSD from from two two independent independent experiments experiments ± carriedcarried out out in triplicate;in triplicate; the datathe data for experiment for experime (c)nt show (c) meanshow valuesmean fromvalues two from independent two independent repetitions carriedrepetitions out in carried duplicate; out in for duplicate; better readability, for better readability, SDs were not SDs included. were not included.

Next, using time-kill assay we showed that after 24 h of incubation, the combinations of FLC1-PB8 and FLC1-PB16 (µg/mL) are fungicidal, which was expressed by one order and by two orders of Int. J. Mol. Sci. 2019, 20, 5577 6 of 21 decrease in viability, respectively (Figure3b). In the presence of OMVs, this fungicidal action of both drugs was significantly abolished or weakened in comparison to action of a single drug and depending on PB concentration used. This protective effect of Mc6 OMVs against C. albicans was also confirmed in bacterial growth kinetics (Figure3c). By estimating the profiles of Candida growth, as a result of every hour measurements carried out for 24 h at 37 ◦C, we have documented that the OMVs added to both aforementioned drug combinations abolished their inhibitory effect on Candida growth rate. Thus, the Candida growth in the presence of OMVs remained at the level for a single compound. Collectively, these results indicate that OMVs protect C. albicans from PB-dependent fungicidal effect in combination with fluconazole.

2.5. Mc6 OMVs Passively Enhance the Virulence of Pathogenic Yeasts Facilitating the Formation of Filaments It was documented that azole drugs are necessary both during growth and induction step to have an effect on transition yeast-to hyphae in C. albicans [23]. The presence of FLC in the decreased range between 1–0.125 µg/mL both during growth and induction step caused considerable inhibition of hyphal growth (data not shown). Therefore, to determine the effects of fluconazole in combination with polymyxin B on hyphal growth, the yeasts were initially preincubacted overnight at 37 ◦C with 1/16 MIC of FLC (0.0625 µg/mL). During the 2 h induction of hyphal growth in the presence of 10% fetal bovine serum, 1/16 MIC of FLC was added together with 1/8, 1/16, and 1/32 MIC of PB in the presence of absence of 20 µg/mL of OMVs. Figure4a,b shows that the number of hyphal cells, shown by the ratio of cells in yeast form compared to that with formed filaments after 2 h of incubation, decreases as the concentration of polymyxin B increases. Under the tested conditions, the combinations FLC0.0625-PB8 and FLC0.0625-PB16 (µg/mL) showed almost complete inhibitory activity against the formation of hyphae. FLC0.0625-PB4 had also clear inhibitory effect on yeast-to-hyphae transition in comparison to FLC alone. The presence of 20 µg/mL OMVs in all tested drug combinations caused a significant increase in the percentage of hyphae-forming cells (Figure4b), probably as a result of neutralizing PB-dependent potentiating effect of FLC. Furthermore, the increase in yeast-to-hyphal transition was accompanied by a significant extension of filaments (Figure4c). In summary, these results indicate that in the presence of Mc6 OMVs, the inhibiting effect of FLC and PB, at concentrations significantly lower than MICs, on the formation of filaments is abolished. It shows that in the presence of OMVs, C. albicans may retain the virulent filamentous phenotype in the presence of both antifungal drugs, thus increasing its virulence.

2.6. OMV-Dependent Protection against PB and Complement is the Result of PB Sequestration on Free OMVs To study how efficiently PB is neutralized by Mc6 OMVs, various biological and biochemical methods were used. Initially, a decrease of free PB was confirmed indirectly using E.coli ML35p mutant with constitutive β-galactosidase expression (Figure5a). In the experiment, different concentrations of Mc6 OMVs were introduced into the system containing E. coli ML-35p mutant at a concentration of ~106 cfu/mL and PB at a concentration of 5 µg/mL. By measuring the quantitative intracellular influx and hydrolysis of ONPG (β-galactosidase substrate), as a result of membrane permeabilization, we showed that OMV-dependent protection was dose-dependent and that PB was very quickly depleted from the environment in the presence of at least 20 µg/mL OMVs leading to a decrease in peptide activity (Figure5a). Int. J. Mol. Sci. 2019, 20, x FOR PEER REVIEW 6 of 20

Next, using time-kill assay we showed that after 24 h of incubation, the combinations of FLC1- PB8 and FLC1-PB16 (μg/mL) are fungicidal, which was expressed by one order and by two orders of decrease in viability, respectively (Figure 3b). In the presence of OMVs, this fungicidal action of both drugs was significantly abolished or weakened in comparison to action of a single drug and depending on PB concentration used. This protective effect of Mc6 OMVs against C. albicans was also confirmed in bacterial growth kinetics (Figure 3c). By estimating the profiles of Candida growth, as a result of every hour measurements carried out for 24 h at 37 °C, we have documented that the OMVs added to both aforementioned drug combinations abolished their inhibitory effect on Candida growth rate. Thus, the Candida growth in the presence of OMVs remained at the level for a single compound.

Int.Collectively, J. Mol. Sci. 2019 these, 20, 5577results indicate that OMVs protect C. albicans from PB-dependent fungicidal effect7 of 21 in combination with fluconazole.

FigureFigure 4. 4.Eff Effectect of OMVsof OMVs on filamentouson filamentous growth growth of C. of albicans C. albicans: Yeast: cellsYeast were cells preincubated were preincubated overnight withovernight 1/16 MIC with of FLC1/16 (0.0625MIC ofµ FLCg/mL) (0.0625 without μg/mL) shaking without and then,shaking after and washing, then, after were washing, incubated were at OD = 0.4incubated in 0.4 mL at OD of hyphae = 0.4 in inducing0.4 mL ofmedium hyphae inducing (10% FBS medium in PBS) in(10% 24-well FBS in microtiter PBS) in 24-well plates withmicrotiter shaking

(130plates rpm) with for 2shaking h at 37 ◦(130C in therpm) presence for 2 h ofat 0.0625 37 °Cµ ing/ mLthe FLC,presence indicated of 0.0625 concentrations μg/mL FLC, of PB,indicated and with orconcentrations without 20 µg /ofmL PB, of and OMVs. with Threeor without independent 20 μg/mL experiments of OMVs. Three were inde performed.pendent (experimentsa) The filamentation were wasperformed. monitored (a under) The invertedfilamentation microscope was monitored using 40 underobjective inverted (Zeiss) microscope and images using were 40× recorded. objective Scale × bars(Zeiss)= 100 andµm. images (b) The were number recorded. of hyphal Scale cellsbars versus= 100 μ yeastm. (b) cells The wasnumber determined of hyphal using cells ImageJ versus software. yeast Thecells pool was of determined yeast cells using contains ImageJ yeasts software. and yeasts The pool whose of yeast germ cells tubes contains did not yeasts exceed and 10 yeastsµm. whose The data representgerm tubes the did mean not values exceed calculated 10 μm. The for data200 represent cells for the each mean of values the eight calculated tested options. for ≥ 200 ( ccells) Box for and each of the eight tested options. (c) Box and≥ Whisker plots of hyphal length: Minimal and maximal Whisker plots of hyphal length: Minimal and maximal values, median, quartiles Q1 and Q3. Statistical values, median, quartiles Q and Q . Statistical analysis was performed by one-way ANOVA (*p < analysis was performed by one-way1 3 ANOVA (* p < 0.05, ** p < 0.01). 0.05, ** p < 0.01). Next, to directly confirm the role of OMVs in PB sequestration, we determined zeta potential, showing2.5. Mc6 that OMVs the Passively initially Enhance moderately the Virulence negative of zetaPathogenic potential Yeasts of Facilit intactating OMVs the Formation was immediately of and significantlyFilaments neutralized by addition of at least 50 µg/mL of PB, in a concentration-dependent and diluent-dependent manner leading to membrane depolarization (Table1). The presented results of alteration in Zeta potentials were similar after 30 min and 60 min incubation at 37 ◦C (data not shown). Finally, to quantitatively assess the magnitude of PB sequestration by OMVs, the residual free PB content remaining after ultrafiltration of OMV-PB complexes was measured using HPLC-UV. Figure6 shows the chromatograms of recovered by ultrafiltration PB that was preincubated for 30 min either alone or with OMVs. The loss of PB following incubation with OMVs at 5 µg/mL and 20 µg/mL was almost 60% (p < 0.001) and over 96% (p < 0.001), respectively, in reference to standard PB (Table2). Int. J. Mol. Sci. 2019, 20, x FOR PEER REVIEW 7 of 20

It was documented that azole drugs are necessary both during growth and induction step to have an effect on transition yeast-to hyphae in C. albicans [23]. The presence of FLC in the decreased range between 1–0.125 μg/mL both during growth and induction step caused considerable inhibition of hyphal growth (data not shown). Therefore, to determine the effects of fluconazole in combination with polymyxin B on hyphal growth, the yeasts were initially preincubacted overnight at 37 °C with 1/16 MIC of FLC (0.0625 μg/mL). During the 2 h induction of hyphal growth in the presence of 10% fetal bovine serum, 1/16 MIC of FLC was added together with 1/8, 1/16, and 1/32 MIC of PB in the presence of absence of 20 μg/mL of OMVs. Figure 4a,b shows that the number of hyphal cells, shown by the ratio of cells in yeast form compared to that with formed filaments after 2 h of incubation, decreases as the concentration of polymyxin B increases. Under the tested conditions, the combinations FLC0.0625-PB8 and FLC0.0625-PB16 (μg/mL) showed almost complete inhibitory activity against the formation of hyphae. FLC0.0625-PB4 had also clear inhibitory effect on yeast-to-hyphae transition in comparison to FLC alone. The presence of 20 μg/mL OMVs in all tested drug combinations caused a significant increase in the percentage of hyphae-forming cells (Figure 4b), probably as a result of neutralizing PB-dependent potentiating effect of FLC. Furthermore, the increase in yeast-to-hyphal transition was accompanied by a significant extension of filaments (Figure 4c). In summary, these results indicate that in the presence of Mc6 OMVs, the inhibiting effect of FLC and PB, at concentrations significantly lower than MICs, on the formation of filaments is abolished. It shows that in the presence of OMVs, C. albicans may retain the virulent filamentous phenotype in the presence of both antifungal drugs, thus increasing its virulence.

2.6. OMV-Dependent Protection against PB and Complement is the Result of PB Sequestration on Free OMVs To study how efficiently PB is neutralized by Mc6 OMVs, various biological and biochemical

Int.methods J. Mol. Sci. were2019 , used.20, 5577 Initially, a decrease of free PB was confirmed indirectly using E.coli ML35p8 of 21 mutant with constitutive β-galactosidase expression (Figure 5a).

FigureFigure 5. 5.OMVs OMVs block the the permeabilizing permeabilizing activity activity of ofmembrane membrane targeting targeting agents agents.. Change Change in E.coli in E.ML- coli ML-35p35p membrane membrane permeability permeability was was assayed assayed by a by time-dependent a time-dependent influx influx of ONPG of ONPG in the in presence the presence of ofmembrane-targeting membrane-targeting agents agents and and diffe direntfferent concentrations concentrations of Mc6 of Mc6OMVs OMVs.. Bacteria Bacteria at log phase at log ( phase∼106 6 (~10cfu/mL)cfu /weremL) incubated were incubated on microplate on microplate at 37 °C for at 120 37 min◦C forin the 120 presence min in of the indicated presence concentrations of indicated concentrationsof OMVs together of OMVs with together(a) polymyxin with ( aB) (PB) polymyxin and (b) B normal (PB) and human (b) normal serum human (NHS). serum The absorbance (NHS). The absorbancewas measured was measuredat indicated at indicatedtime points. time The points. result Thes are results shown are as shown mean as± SD mean fromSD at fromleast atthree least ± threeindependent independent experiments experiments performed performed in duplicate. in duplicate.

In the experiment, different concentrations of Mc6 OMVs were introduced into the system containing E. coliTable ML-35p 1. Alteration mutant at in Zetaa concentration potential of OMVsof ∼10 treated6 cfu/mL with and polymyxin PB at a B.concentration of 5 μg/mL.PB By Bindingmeasuring by OMVs the quantitative in NaPB Bu ffintracellularer influx and PB hydrolysis Binding by of OMVs ONPG in ( miliQβ-galactosidase substrate), as a result of membrane permeabilization, we showed that OMV-dependent protection Zeta Potential Zeta Potential Treatment Treatment (mV) (mV) 20 µg/mL OMVs (control) 22.5 1.05 20 µg/mL OMVs (control) 24.7 0.76 − ± − ± 20 µg/mL OMVs + 5 µg/mL PB 21.2 1.06 20 µg/mL OMVs + 5 µg/mL PB 25.9 0.97 − ± − ± 20 µg/mL OMVs + 50 µg/mL PB 16.5 0.76 * 20 µg/mL OMVs + 50 µg/mL PB 9.6 0.35 * − ± − ± 20 µg/mL OMVs + 250 µg/mL PB 10.1 0.46 * 20 µg/mL OMVs + 250 µg/mL PB 1.17 0.14 * − ± − ± *-p < 0.001 in reference to control as determined by one-way ANOVA.

Collectively, our results demonstrate that PB is quickly and effectively depleted from the environment via OMV-dependent sequestration. Similar experiments with E.coli ML35p mutant were performed with human active serum (NHS). Similar to PB, the strong permeabilizing potency of membrane attack complex (MAC), present in 10% human serum, against E.coli ML35p was considerably decreased or even dumped as the concentration of OMVs increased (Figure5b). It suggests that by deposition of complement components on the OMV surface, vesicles trigger MAC formation away from target bacteria and thus protect them from MAC-mediated lysis. Accordingly, based on the quantitative formation of soluble non-proteolytic membrane attack complex (SC5b-9), we demonstrated that the Mc6 OMVs activate human serum complement in a concentration-dependent manner (Figure S1). Int. J. Mol. Sci. 2019, 20, x FOR PEER REVIEW 8 of 20 was dose-dependent and that PB was very quickly depleted from the environment in the presence of at least 20 μg/mL OMVs leading to a decrease in peptide activity (Figure 5a).

Table 1. Alteration in Zeta potential of OMVs treated with polymyxin B.

PB Binding by OMVs in NaPB Buffer PB Binding by OMVs in miliQ Zeta Potential Zeta Potential Treatment Treatment (mV) (mV) 20 μg/mL OMVs (control) −22.5 ± 1.05 20 μg/mL OMVs (control) −24.7 ± 0.76 20 μg/mL OMVs + 5 μg/mL PB −21.2 ± 1.06 20 μg/mL OMVs + 5 μg/mL PB −25.9 ± 0.97 20 μg/mL OMVs + 50 μg/mL PB −16.5 ± 0.76* 20 μg/mL OMVs + 50 μg/mL PB −9.6 ± 0.35* 20 μg/mL OMVs + 250 μg/mL PB −10.1 ± 0.46* 20 μg/mL OMVs + 250 μg/mL PB −1.17 ± 0.14* Int. J. Mol. Sci. 2019, 20, 5577 9 of 21 *- p < 0.001 in reference to control as determined by one-way ANOVA.

(a)

PB2

PB1

(b)

PB2

PB1

(c)

PB2

PB1

FigureFigure 6. The 6. TheHPLC HPLC spectra spectra of ofpolymyxin polymyxin B B incubated incubated alone oror along along with with OMVs. OMVs. Chromatograms Chromatograms of (a)of 50 ( aμ)g/mL 50 µg/ mLsolution solution of polymyxin of polymyxin B, B, (b (b) )50 50 μµg/mLg/mL of PBPB afterafter treatment treatment with with 5 µ5g /μmLg/mL of Mc6of Mc6 µ µ OMVs,OMVs, (c) 50 (c μ) 50g/mLg/ mLof PB of PBafter after treatment treatment with with 20 20 μg/mLg/mL ofof Mc6Mc6 OMVs. OMVs. Samples Samples were were prepared prepared as as described in Materials and Method section. HPLC analyses were performed with Macherey–Nagel described in Materials and Method section. HPLC analyses were performed with Macherey–Nagel Nucleodur C18 Isis column. UV-detection at 215 nm. Double peak with Tr = 3.8 min corresponds Nucleodur C18 Isis column. UV-detection at 215 nm. Double peak with Tr = 3.8 min corresponds mainly to polymyxin PB2 and PB3, and double peak with Tr= 7 min corresponds to polymyxin PB1 and PB1-I. The representative chromatograms are shown. Int. J. Mol. Sci. 2019, 20, 5577 10 of 21

Table 2. Magnitude of PB sequestration on M. catarrhalis OMVs.

Treatment PB2 (µg/mL) PB1 (µg/mL) PB2 + PB1 (µg/mL) Standard PB 14.38 0.48 17.45 3.58 31.83 3.47 ± ± ± PB after treatment with 5 µg/mL OMVs 7.76 1.01 5.19 1.05 13.67 1.33 * ± ± ± PB after treatment with 20 µg/mL OMVs 0.96 0.42 0.06 0.01 1.02 0.42 * ± ± ± The PB content was measured by HPLC-UV method as described in material and methods. The results are expressed as mean SD. * p < 0.001 in reference to standard PB as determined by one-way ANOVA. ±

2.7. OMVs Associated with Bacteria do not Protect against PB To answer the question of whether Mc6 OMVs protect bacteria against the membrane-active agent by sequestration only indirectly, being far from the bacterial surface or through the direct shield of these cells, we determined the association capability of OMVs using flow cytometry. We used 4-times higher concentration of OMVs (80 µg/mL) to make sure that the entire cell population could be covered (Figure S2). We found that Mc6 OMVs incubated with studied strains strongly interacted within 30 min with the parental strain as well as with other strains from Moraxella species. After this time, more than 93% of measured events were fluorescent, indicating that FITC-labelled OMVs were associated with unlabeled bacteria. No increase in fluorescence was observed when Mc6 OMVs were incubated even up to 2 h with bacteria from other species indicating the complete lack of association (Figure7a,b). These results indicate that OMVs released by M. catarrhalis associated only in intraspecies but not in interspecies tested systems, pointing to the specificity of this reaction. To verify whether the covering of M. catarrhalis with OMVs is another mode of protection against PB, 4 h time-kill assays (Figure7c,d), as well as 24 h real-time bacterial growth kinetics (Figure7e,f) were carried out on OMV-associated and not associated cells in the presence and absence of PB. Not in the line with expectation, the results showed that for M. catarrhalis associated and not associated with OMVs, the lethal effect of PB occurred after 4 h (Figure7c) and 2 h (Figure7d), respectively. It indicates that the association does not ensure longer protection but only delays the bactericidal effect. The lack of protection was confirmed in growth kinetic experiments documenting that PB-dependent growth inhibition for OMV-associated M. catarrhalis and non-associated control was comparable and preserved for 24 h of incubation (Figure7e). In the case of bacteria incubated in the presence of free OMVs and PB, the OMV-dependent protection is assured for ~7 h being on the level of control growth. Some modest and stable level of protection is maintained for 24 h (Figure7f). Compared to the growth profiles of OMV-associated and non-associated M. catarrhalis, it was shown that the interaction with vesicles weakens the growth dynamics of the former. Overall, the results are evidence that OMVs associated with bacteria do not protect against PB, assuming of course that the bacterial cell is completely shielded by them. Consequently, it suggests the active fusion between OMVs and target cell membrane rather than only surface association. Int. J. Mol. Sci. 2019, 20, x FOR PEER REVIEW 10 of 20

was comparable and preserved for 24 h of incubation (Figure 7e). In the case of bacteria incubated in the presence of free OMVs and PB, the OMV-dependent protection is assured for ∼7 h being on the level of control growth. Some modest and stable level of protection is maintained for 24 h (Figure 7f). Compared to the growth profiles of OMV-associated and non-associated M. catarrhalis, it was shown that the interaction with vesicles weakens the growth dynamics of the former. Overall, the results are evidence that OMVs associated with bacteria do not protect against PB, assuming of course that the Int. J. Mol.bacterial Sci. 2019 cell, 20 is, 5577completely shielded by them. Consequently, it suggests the active fusion between11 of 21 OMVs and target cell membrane rather than only surface association.

FigureFigure 7. Mc6 7. Mc6 OMVs OMVs highly highly associated associated with with bacteria bacteria do not protect protect against against polymyxin polymyxin B-dependent B-dependent killing.killing. (a) Flow (a) Flow cytometry cytometry analysis analysis of of fluorescein fluorescein isothiocyanateisothiocyanate (FITC)-labelled (FITC)-labelled OMVs OMVs associated associated with intraspecieswith intraspecies bacteria bacteria (upper (upper panel) panel) and and not not associatedassociated with with interspecies interspecies bacteria bacteria (lower (lower panel), panel), The fluorescence intensities of OMV-associated bacteria are shown as black histograms whereas The fluorescence intensities of OMV-associated bacteria are shown as black histograms whereas control control bacteria as dotted histograms. Representative plots are shown. (b) Quantification of bacteria bacteria as dotted histograms. Representative plots are shown. (b) Quantification of bacteria associated associated with FITC-labelled OMVs. Data are expressed as mean fluorescent intensity (MFI) ± SD with FITC-labelled OMVs. Data are expressed as mean fluorescent intensity (MFI) SD from three ± independent experiments performed in duplicates. (c) Bactericidal activity of polymyxin B against M. catarrhalis after association with OMVs. (d) Bactericidal activity of polymyxin B against M. catarrhalis incubated with free OMVs. (e) 24 h growth-curve kinetics for M. catarrhalis after association with OMVs. (f) 24 h growth-curve kinetics for M. catarrhalis incubated with free OMVs. The kinetics was measured using Varioskan™ LUX reader with measurements at 30 min intervals. In experiments (c,d) results are shown as mean SD from at least two independent experiments. In experiments (e,f) results are ± shown as mean from at least three independent experiments performed at duplicate and for better readability, SDs were not included. Int. J. Mol. Sci. 2019, 20, 5577 12 of 21

3. Discussion Bacteria and yeasts employ a variety of means to protect their envelopes against harmful environmental factors. Some of these factors cause membrane permeabilization, while others can inhibit the synthesis of cellular membrane components affecting the physical properties of the membrane [24,25]. The common strategy of bacteria to avoid membrane targeting factors is to overcome the negative charges of their surface envelope [24,26]. Alternatively, bacteria can release OMVs that act as extracellular decoys for some antimicrobials. In this study, we report that OMVs may serve as a model of passive interspecies protection of prominent pathogenic bacteria but also pathogenic yeasts. While documenting the lethal effect, we showed that free OMVs block both the bactericidal action of peptide antibiotic polymyxin B and the lytic activity of complement. By quantitative measurement of alteration in electric charge on the OMV surface treated with PB as well as free PB determination by HPLC, we showed that 20 µg/mL of OMVs that had a significant protective effect against interspecies microorganisms caused almost complete depletion of this model cationic antimicrobial agent after co-incubation, thus indicating the immediate sequestration of the peptide on free OMVs. Accordingly, in similar experiments, we showed that OMVs inhibit serum lytic activity against prone pathogens, indicating that by deposition of complement components on the OMV surface, vesicles trigger MAC (SC5b-9) formation away from target bacteria and thus protect them from MAC-mediated lysis. The intensity of OMV-dependent complement activation in vitro was correlated with the number of free vesicles in the environment. Therefore, it is possible that any fluctuation in the number of released vesicles may positively or negatively influence complement activation of this potent innate mechanism. Next, using the PB-dependent model, we investigated whether OMVs-dependent protection is a signature of only free vesicles or those associated with the cell. Using flow cytometry, we documented that the association of OMVs with bacterial cells may be very potent and species-specific, influencing the cell sensitivity to antimicrobial compound. Our vesicles were able to associate only with representatives of M. catarrhalis species while they did not work in interspecies systems. By evidencing the lethal effect of cell-associated OMVs exposed to polymyxin B, we showed that OMVs do not protect against AMPs this way, suggesting a fusion between OMVs and OM rather than only shielding. It is tempting to speculate that the specificity of interaction between OMVs and recipient cell is somehow pivotal in its sensitization at least to AMPs. We are therefore in agreement with Tashiro et al. that elucidating the selectivity in OMV-cell interaction is critical for an improved understanding of the outcome of this reaction. On the other hand, it has been documented that interactions with OMVs for some cross-pathogens can be specific, less specific, or not specific at all [27]. Thus, the OMV-dependent interplay may contribute to the generation of synergistic or antagonistic interactions between pathogens, resulting in more or less harmful outcomes for the host. Next, we address the question of whether OMV-dependent trapping of cationic antimicrobials can protect pathogens from other kingdoms such as fungi of the species C. albicans. Polymyxins alone are effective against C. albicans only at relatively high concentrations [28]. Azoles including fluconazole are common antifungal drugs [29]. They affect the integrity of fungal membranes, altering their morphology and inhibiting growth [30,31]. It has been reported that quinolone and other antibiotics may augment the anti-candidal activity of azole and polyene agents [32]. Previous research also documented that PB in lower concentrations exerts a potent antifungal effect when combined with fluconazole [22,33]. Therefore, the elimination of the antibiotic from these systems by means of OMVs should decrease the fungus susceptibility to fluconazole. To test this hypothesis, we incubated C. albicans in the presence of both drugs and OMVs using FLC at MIC50 (1 µg/mL) and PB in the range much lower than MIC (1/8–1/16). To our knowledge, for the first time, we have provided evidence that bacterial OMVs can inhibit the fungicidal action of certain combined antifungal agents that are effective against pathogenic yeasts. Next, we documented that sub-MIC concentrations of FLC (1/16) and PB (in the range of 1/8–1/32) applied together weakened or almost completely inhibited the yeast-to-hyphae transition. Furthermore, in the presence of OMVs, the filamentous phenotype was Int. J. Mol. Sci. 2019, 20, 5577 13 of 21 considerably recovered and even strengthened despite exposure to both drugs. It is a very undesirable action of OMVs, since filamentation increases the virulence of C. albicans, which in hyphae form is more invasive and can attach in a higher number to epithelial cells than yeast and pseudohyphae forms [34,35]. Using analogies to the role of OMVs contributing to the sequestration of PB, it is tempting to speculate that the same exists in this case. Collectively, both of the aforementioned activities of bacterial vesicles seem to render C. albicans less vulnerable to destruction. The documented inter-kingdom OMV-based mutualistic relationship between bacteria and yeasts are, to our knowledge, a novel phenomenon. Therefore, its importance for more complex in vitro and in vivo conditions, as well as pathophysiology, remains to be clarified. The complex Candida–bacteria interactions are not a rare occurrence and may have an important impact on the human disease by causing e.g., the faster biofilm growth or Candida-dependent induction of antimicrobial resistance of Staphylococci [36,37]. Sometimes, the results investigating cross-kingdom polymicrobial interactions are very contradictory even for the same microbial components. It was shown that direct or indirect (by released soluble molecules) contact of P. aeruginosa or A. baumannii with C. albicans or other fungi may lead to the killing of yeast [38,39]. It can also decrease fungal filamentation, biofilm formation, and conidia biomass [40]. The synergistic collaboration between the two pathogens was also documented. For example, the pre-colonization with C. albicans, which compromises the immune system, facilitates the emergence of A. baumannii pneumonia [41]. The interactions between microbes are not only affected by the specific combination of microorganisms, but also by the environment such as immunological milieu. Therefore, bacterial OMVs shape the behavior of neighboring microbes and the overall outcome of their interplay for the host. Due to the limited antifungal arsenal, the synergistic effect of PB and fluconazole or human broad-spectrum AMP lactoferrin with amphotericin B or fluconazole, which increases the activity of the antifungals against Candida spp, could be an alternative for treatment [22,42]. Likewise, AMPs such as defensins or gramicidin have shown to be a promising alternative to the current antimycotic and antibacterial therapies [43–45]. Furthermore, AMPs display a lower propensity to develop resistance than do conventional antibiotics [45]. In light of our research, however, these promising strategies should consider the unfavorable role of OMVs as a trap for host cationic peptides in mixed bacterial and bacterial–fungal infections. Our results may also have a meaning in medical microbiology. Because AMPs are already used in topical nasal antimicrobials in the treatment of nasal or paranasal cavity infections (sinusitis, maxillary, otitis media) [46,47], it is conceivable that abundantly produced OMVs, by very efficient AMP sequestration, may decrease the pharmacokinetics of these compounds. Another important issue of our results is the number of OMVs needed to show biological activity. In general, the information about the number of OMVs produced in vivo is still very limited. Although OMV production in the course of infection has been documented [48], the magnitude of this production was not given. During the in vitro part of this study, however, the biologically active concentration of OMVs was 5 µg/mL per 103 cfu/mL. In our study, 20 µg/mL of OMVs was protective against 106 cfu/mL, indicating that OMV amounts used in our study are rational and may resemble the infectious condition.Furthermore, there is many data on how different stress factors may induce bacterial hypervesiculation incuding temperature stress [49], oxidative stress [50], hyperosmotic stress [49], or antibiotic stress [51]. So far, the correlation between OMVs production and pathophysiology of a specific disease was proven both in animal sepsis-like inflammation models [52–54] and in a patient with fatal meningococcal septicaemia [55]. Based on these aforementioned examples, the OMVs concentrations, which seems to be clinically relevant, are in the range 5–20 µg/mL. Overall, our results on the protective and somehow deleterious for pathogens role of OMVs, in the bacteria–bacteria and bacteria–yeast interspecies systems, underline the enormous potential of these nanostructures as accelerating factors in case of various mixed infections. Furthermore, the OMV-dependent mode of actions may serve as a model of passive resistance of gram-negative bacteria not only to antimicrobials, but also to humoral defense components, which operate to disrupt cell membrane. Likewise, for dimorphic yeasts, the ability of OMVs to sequester membrane active Int. J. Mol. Sci. 2019, 20, 5577 14 of 21 compounds that augment the antifungal activity of azoles may have an important impact on Candida virulence. This work may serve as an important basis for further evaluation of OMVs-dependent interactions within pathogenic bacterial-fungal communities. Our results indicate that OMVs are important players in interspecies and cross-kingdom microbial interactions.

4. Materials and Methods

4.1. Materials

4.1.1. Reagents BHI (Brain Heart Infusion, OXOID, Basingstoke, UK) ); TSB (Tryptone Soya Broth, OXOID, Hampshire, England); TSA (Tryptone Soya Agar, OXOID, Hampshire, England), Bradford reagent (Protein Assay Dye Reagent Concentrate, Bio-Rad, München, Germany); β-nicotinoamide adenine dinucleotide hydrate (NAD, Sigma-Aldrich, Steinheim, Germany); fluconazole (FLC, Sigma-Aldrich, Poznan, Poland), hemin (Sigma-Aldrich, St. Louis, MO, USA); polymyxin B sulfate salt (PB, Sigma-Aldrich, Denmark); fluorescein isothiocyanate (FITC, ThermoScientific, Rockford, IL, USA); Hank’s Buffer with Ca2+, Mg2+ (HBSS, PAN Biotech, UK); RPMI 1640 (Lonza, Walkersville, MD, USA); o-nitrophenyl-β-D-galactopyranoside (ONPG, Sigma, Steinheim Germany), heat inactivated (56 ◦C, 1 h) FBS (Fetal bovine serum, Gibco Life Technologies, Grand Island, NY, USA).). HPLC chemicals: Acetonitrile (Sigma, München, Germany) for separation was HPLC far UV/gradient grade (J. T. Baker, Avantor™ Performance Material); 32 mM Na2SO4 solution for chromatographic usage was prepared with 4.5 g anhydrous sodium sulfate (POCh, Avantor™ Performance Material, Gliwice, Poland) and MiliQ (ultrapure water made with Simplicity UV Water Purification System, Merck Millipore, Saint-Quentin, France).

4.1.2. Microbial Strains and Growth Condition The following microbial strains were used: Moraxella catarrhalis (Mc5, Mc6, Mc8), nontypeable Haemophilus influenzae (NTHi3, NTHi6), Acinetobacter baumannii (ATCC 19606), Pseudomonas aeruginosa (PAO1), Candida albicans (Ca1), mutant of ML-35p, a lactose permease-deficient strain with constitutive cytoplasmic β-galactosidase. All strains were from the collection of our Institute. M. catarrhalis strains were grown on Columbia agar plates or BHI broth. NTHi strains were grown on chocolate agar plates or in BHI broth supplemented with hemin and NAD at final concentrations of 15 µg/mL each. Moraxella and Haemophilus strains were cultivated at 37 ◦C with 5% CO2. A. baumannii; E. coli ML-35p and P. aeruginosa were routinely cultured in TSB medium at 37 ◦C. C. albicans was cultured in yeast extract-peptone-glucose (YPG) in 37 ◦C.

4.2. Methods

4.2.1. Outer Membrane Vesicles Isolation Outer membrane vesicles (OMVs) isolation was performed as described previously [21]. The protein concentartions of purified OMVs preparations was determined by Bradford assay) and the quality of OMVs preparation was confirmed in 12% SDS-PAGE.

4.2.2. Time-Kill Assay For testing PB or human serum (NHS) activity, the log-phase bacterial suspension (5 105–106 cfu/mL) was incubated with or without PB (in the range 0.5–5 µg/mL) or NHS (in the × range 25%–75%) with the presence or absence of free OMVs (2 µg/mL or 20 µg/mL) in the final volume of 200 µL 1% medium (w/v). The experiments were performed from 0 to 240 min and the 10 µL aliquots of 10-time diluted bacterial suspensions were plated in triplicate on appropriate agar plates at 0, 30, 120, and 240 min time points. The colony counts and cfu/mL were calculated the next day. The controls for Int. J. Mol. Sci. 2019, 20, 5577 15 of 21

NHS contained heat-inactivated serum (HiNHS), (56 ◦C, 30 min). The bactericidal activity of PB or NHS was expressed in each time point as cfu/mL in reference to cfu/mL in time 0. Analogous experiments with bacteria associated and non-associated with OMVs (20 µg/mL or 80 µg/mL) in the presence of PB (5 µg/mL) were performed. Analogous experiments for selected combinations of antifungals were carried out for C. albicans except that: (i) Initial inoculum was ~2 105–5 105 cfu/mL, (ii) tested × × antimicrobials were PB (8 µg/mL or 16 µg/mL) and fluconazole (1 µg/mL), (iii) the medium was 0.5% YPG (w/v), and (iv) incubation lasts 24 h. All microbicidal assays were performed at least two times in triplicate.

4.2.3. Growth Kinetics Assay All growth kinetics experiments were performed on the flat-bottomed 96-well microplates (NUNC, Denmark) at 37 ◦C in volume 200 µL. Dynamic of growth was measured using Varioskan™ LUX reader with measurements at 30 min intervals for bacteria at 106 cfu/mL diluted (final concentration) in 1% BHI (w/v) and 60 min intervals for yeast at 105 cfu/mL (final concentration.) diluted in 0.5% YPG. OMV-associated bacteria preparation: 0.5 mL of ~108 cfu/mL exponentially growing bacteria were washed by centrifugation with HBSS Ca2+ Mg2+ and resuspended in 100 µL OMVs (20 µg/mL or 80 µg/mL) in Eppendorf tube. Association was performed for 30 min at 37 ◦C with gentle mixing. 2+ 2 Thereafter, all samples were washed with HBSS Ca , Mg by centrifugation (8000 rpm, 10 min, 4 ◦C) to remove free OMVs particles, diluted to 2 106 cfu/mL with HBSS Ca2+, Mg2, and used in growth × kinetics assay.

4.2.4. Checkerboard Microdilution Assay The microdilution assay was performed on flat-bottom microplate according to the CLSI (formerly NCCLS) standard [56] except that the initial inoculum for C. albicans was ~105 cfu/mL and cells were incubated at 37 ◦C without shaking. Synergy/growth potentiation was tested by the checkerboard method including a two-dimensional array of serial concentrations of both drugs. The fluconazole was used in concentrations 1–64 µg/mL whereas polymyxin B in concentrations 1–128 µg/mL. Wells without drugs or yeast inoculation were included as positive and negative controls, respectively. The MIC100 of polymyxin B and the MIC50 of fluconazole was defined as the lowest drug concentration that caused a decrease in absorbance of 100% and 50%, respectively, compared to control in drug-free medium.

4.2.5. Induction of Filamentation Before the induction of filamentation, C. albicans cells were preincubated overnight in YPG supplemented with sub-MIC concentration of FLC at 37 ◦C without shaking. To induce hyphal transition, the suspensions of C. albicans (OD = 0.4) were treated with 10% heat inactivated FBS in PBS for 2 h at 37 ◦C in 24-well flat-bottom microplates (NUNC) in volume 0.4 mL, in the presence of sub-MIC concentrations of FLC (1/16) and PB (range 1/8–1/32) and with shaking (130 rpm). The samples were observed under inverted microscope Zeiss Axio Vert. A1 with objective Zeiss LD A-Plan (40 /0.55 Ph1). × The images wer recorded using Industrial Digital Camera 5.1 MP 1 /2.5”. The assessment of cell morphology and the length (µm) of hyphae was performed using ImageJ software.

4.2.6. Membrane Permeability Assay with ONPG To assess the polymyxin B sequestration by M. catarrhalis OMVs in vitro, the time-dependent decrease of permeabilizing activity of this peptide against E. coli ML-35p with constitutive β-galactosidase activity was measured using the ONPG-mediated β-galactosidase microplate assay as previously described [57]. The final bacterial suspension (~106 cfu/mL) in 10 mM sodium phosphate NaPB (pH 7.4) were incubated in the flat-bottom 96-well microplates (NUNC, Denmark) with 5 µg/mL of PB in the presence of OMVs at concentration range 1–20 µg/mL and with 3 mM ONPG as β-galactosidase substrate in the final volume of 150 µL. Microplates were incubated at 37 ◦C for 1.5 h and optical Int. J. Mol. Sci. 2019, 20, 5577 16 of 21 densities at λ = 405 nm were measured every 15 min (spectrophotometer ASYS). All the assays were performed at least 3 times in duplicates.

4.2.7. Complement Activation This method was described previously [58]. To assess complement activation by M. catarrhalis OMVs in vitro, the various concentrations of OMVs (1–20 µg/mL) were pretreated with active normal human serum (NHS) at a volume ratio 1:9 and the soluble terminal complement complex SC5b-9 was measured using ELISA kit according to manufacturer’s instructions (Quidel Corporation, San Diego, USA). Unstimulated NHS served as negative controls.

4.2.8. OMV Association Assay and Flow Cytometry OMVs labeling: Initially, OMVs (80 µg/mL) in PBS were concentrated using 50 kDa Vivaspin centrifugal concentrators (Amicon ultra, Millipore) at 14,000 g for 10 min at 4 C to remove PBS. × ◦ The collected OMVs were reconstituted with 500 µL of 0.05 M carbonate/bicarbonate buffer (pH 9.5) and washed by centrifugation on vivaspin as described before. The collected OMVs at concentration 80 µg/mL were labeled with 1 mg/mL FITC at carbonate/bicarbonate buffer for 30 min at 37 ◦C with gentle mixing in the dark. The remaining fluorochrome was rinsed of 3 times with a 500 µL of cold carbonate/biscarbonate buffer each time, using 50 kDa Vivaspin. The final FITC-labelled OMVs were resuspended in Hank’s buffer Ca2+, Mg2+ at original concentration. OMVs association with bacteria: 1 mL of each fresh bacterial culture corresponding to OD550 = 0.2 was centrifuged and subsequently washed with 1 mL of HBSS Ca2+, Mg2+ (8000 g, 10 min, 4 C). × ◦ The pellet was resuspended in 200 µL of OMV-FITC conjugate (80 µg/mL OMVs) and incubated for 2+ 2 30 min at 37 ◦C with gentle mixing in the dark. Afterward, samples were washed with HBSS Ca Mg by centrifugation (8000 g, 10 min, 4 C) to remove free OMVs particles and finally resuspended in × ◦ 200 µL of HBSS Ca2+ Mg2. Flow cytometric analysis: To detect bacterial cells associated with FITC-labeled OMVs flow cytometry analysis was performed using GUAVA® easyCyte flow cytometer (Millipore, Seattle, WA, USA). Before analysis, the samples were diluted 1:100 to obtain approximately 1–5 105 cell/mL in × HBSS Ca2+ Mg2+. Fluorescence intensity of bacterial cells associated with OMVs was analyzed for green fluorescence in the FL1 channel, by collecting 5000 events. Data were expressed as mean fluorescence intensity (MFI). Data analysis was performed using InCyte Merck Guava software (Millipore, Hayward, CA, USA).

4.2.9. Estimation of Zeta Potential

The Zeta potential of OMVs was measured at room temperature (25 ◦C) by a Zetasizer Nano-ZS 90 (Malvern, UK). The instrument was equipped with a Helium–Neon laser (633 nm) as a source of light. The detection angle of Zetasizer at aqueous media was 173◦. Considering the influence of factors such as conductivity (salt concentration) or pH of the solution on the Zeta potential, to minimize their influence, all Zeta potential measurements were performed with 10 mM NaPB buffer (pH 7.4) or in MiliQ.

4.2.10. HPLC-UV System and Method HPLC chromatography of polymyxin B was performed as described previously [26,59]. The HPLC system consisted of a Water’s 2695 Solvent Manager System with a built-in autosampler and 100 µL sample loop connected to Waters 2996 Photodiode Array Detector. Data collection and peaks integration were realized by computer with Water’s Empower 3 Chromatography Data Software. The separation developed with the use of a Macherey-Nagel EC Nucleodur C18 Isis column (50 mm, 4.6 mm ID with 1.8 µm beads) with Hypersil Gold aQ 5 µm (10 mm 4.6 mm ID) guard precolumn. The mobile phase × consisted of 22% acetonitrile and 78% of 32 mM Na2SO4 in water (pH 3.2 achieved with H2SO4) in isocratic separation mode. Eluent flow 0.75 mL/min and detection realized with a UV detector at Int. J. Mol. Sci. 2019, 20, 5577 17 of 21

215 nm. Column and sample temperature were respectively 30 ◦C and 5 ◦C; separation run time was 15 min. The injection volume was 50 µL. The calibration curve was in the range of 1.56 µg/mL to 100 µg/mL prepared with 10 mg/mL standard stock solution of polymyxin B sulfate salt. Each concentration was injected twice daily for precision into 3 or 2 independent samples, and between day eight samples. The linearity of the method was in a range from 6.25 to 100 µg/mL with 16% of the average coefficient of variation (CV%) for the sum of peaks, and with a coefficient of determination 2 R = 0.997. Since the major constituents of polymyxin B are B1 and B2 [59], (polymyxin B sulfate certificate of analysis, Sigma-Aldrich), these two components were analyzed altogether. To quantify the results, the relative concentration was calculated as the mean PB concentration of sample/mean PB of control 100%. Results were expressed as mean SD. × ± Sample preparation: 200 µL of reaction mixtures containing OMVs (5 µg/mL or 20 µg/mL) and PB (50 µg/mL) or PB alone (50 µg/mL) were incubated for 30 min at 37 ◦C with gentle mixing. The samples were ultrafiltrated using 50 kDa vivaspin centrifugal concentrators (Amicon ultra, Merck Millipore, Cork, Ireland) at 14 000 g for 15 min at RT to remove OMVs. The filtrate was collected and stored at × 20 C until use. Tested and control samples were processed identically. − ◦ 4.2.11. LOS-OMVs Electrophoresis The concentration of lipooligosaccharide (LOS) from Mc6 OMVs was determined based on the purpald assay [60]. Diluted samples were solubilized in Laemmli sample buffer and heated at 100 ◦C 1 for 5 min. Proteinase K (20 mg mL− ) was added per 20 mg of OMV proteins and incubated in a heating block at 60◦C for 1 h. The presence of LOS in proteinase K-treated OMV samples was analyzed by dodecylosulfate gel electrophoresis. The 15 µL of samples were applied to the Glycine-SDS-PAGE (10%) gel path, corresponding to 5 µg LOS. Electrophoresis was carried during 2.5 h at a constant voltage of 80 V at 4 ◦C. The gel was fixed for 1 h at room temperature using a fixing solution (EtOH: Acetic Acid: MiliQ: 40:5:55); POCH, Gliwice, Poland) The fixed gel was stored overnight at 4 ◦C. MiliQ was exchanged for fresh before visualization using a modified Tsai and Frasch [61] silver staining protocol.

4.2.12. TEM The OMVs preparation for TEM was described previously [4]. OMVs were visualized by standard negative staining using a formvar copper grid (Christine Gröpl Electronenmikroskopie, Tulln, Austria) and 2% (w/v) aqueous solution of uranyl acetate. The OMVs were imaged with a TEM operating at an acceleration voltage of 150 kV (Hitachi H-800, Japan).

4.2.13. Statistical Analysis The data were expressed as the mean SD, and analyzed for the significant difference by one-way ± ANOVA or Kruskal–Wallis ANOVA rang using the Statistica (version 13.1) software (StatSoft, Krakow, Poland) Differences were considered statistically significant if p < 0.05.

Supplementary Materials: Supplementary materials can be found at http://www.mdpi.com/1422-0067/20/22/ 5577/s1. Author Contributions: D.A., conceived the study; J.R., P.J., G.G., J.G., DA., performed experiments; D.A., J.R., P.J. and G.G. performed analysis and interpretation of data; A.Z.˙ performed TEM images; D.A. wrote the manuscript with input from J.R.; D.A., J.R., G.G. reviewed and edited the manuscript. D.A., wrote the revised version of the manuscript; Z.D.-K. and D.A. acquired funding. This work is a part of the PhD thesis by J.R. Funding: This study was supported by research grants of University of Wroclaw (1016/S/IGiM/T-20). Conflicts of Interest: The authors declare no conflict of interest. Int. J. Mol. Sci. 2019, 20, 5577 18 of 21

Abbreviations

AMP Antimicrobial peptide FBS Fetal bovine serum FLC Fluconazole HiNHS Heat-inactivated normal human serum LOS Lipooligosaccharide MIC Minimal inhibitory concentration NHS Normal human serum OMV Outer membrane vesicle ONPG o-nitrophenyl-β-D-galactopyranoside PB Polymyxin B

References

1. Guerrero-Mandujano, A.; Hernández-Cortez, C.; Ibarra, J.A.; Castro-Escarpulli, G. The outer membrane vesicles: Secretion system type zero. Traffic 2017, 18, 425–432. [CrossRef][PubMed] 2. Dusane, D.H.; Zinjarde, S.S.; Venugopalan, V.P.; McLean, R.J.C.; Weber, M.M.; Rahman, P.K.S.M. : Implications on rhamnolipid biosurfactant production. Biotechnol. Genet. Eng. Rev. 2010, 27, 159–184. [CrossRef][PubMed] 3. Jan, A.T. Outer Membrane Vesicles (OMVs) of Gram-negative Bacteria: A Perspective Update. Front. Microbiol. 2017, 8, 1053. [CrossRef][PubMed] 4. Augustyniak, D.; Roszkowiak, J.; Wi´sniewska,I.; Skała, J.; Gorczyca, D.; Drulis-Kawa, Z. Neuropeptides SP and CGRP Diminish the Moraxella catarrhalis Outer Membrane Vesicle- (OMV-) Triggered Inflammatory Response of Human A549 Epithelial Cells and Neutrophils. Mediat. Inflamm. 2018, 2018, 1–15. [CrossRef] 5. Bitto, N.J.; Chapman, R.; Pidot, S.; Costin, A.; Lo, C.; Choi, J.; D’Cruze, T.; Reynolds, E.C.; Dashper, S.G.; Turnbull, L.; et al. Bacterial membrane vesicles transport their DNA cargo into host cells. Sci. Rep. 2017, 7. [CrossRef] 6. Kulkarni, H.M.; Nagaraj, R.; Jagannadham, M.V. Protective role of E. coli outer membrane vesicles against antibiotics. Microbiol. Res. 2015, 181, 1–7. [CrossRef] 7. Lee, J.; Lee, E.-Y.; Kim, S.-H.; Kim, D.-K.; Park, K.-S.; Kim, K.P.; Kim, Y.-K.; Roh, T.-Y.; Gho, Y.S. Staphylococcus aureus extracellular vesicles carry biologically active β-lactamase. Antimicrob. Agents Chemother. 2013, 57, 2589–2595. [CrossRef] 8. Manning, A.J.; Kuehn, M.J. Contribution of bacterial outer membrane vesicles to innate bacterial defense. Bmc Microbiol. 2011, 11, 258. [CrossRef] 9. Rumbo, C.; Fernández-Moreira, E.; Merino, M.; Poza, M.; Mendez, J.A.; Soares, N.C.; Mosquera, A.; Chaves, F.; Bou, G. Horizontal transfer of the OXA-24 carbapenemase gene via outer membrane vesicles: A new mechanism of dissemination of carbapenem resistance genes in Acinetobacter baumannii. Antimicrob. Agents Chemother. 2011, 55, 3084–3090. [CrossRef] 10. Chatterjee, S.; Mondal, A.; Mitra, S.; Basu, S. Acinetobacter baumannii transfers the blaNDM-1 gene via outer membrane vesicles. J. Antimicrob. Chemother. 2017, 72, 2201–2207. [CrossRef] 11. Chattopadhyay, M.K.; Jaganandham, M.V. Vesicles-mediated resistance to antibiotics in bacteria. Front. Microbiol. 2015, 6, 758. [CrossRef][PubMed] 12. Kadurugamuwa, J.L.; Mayer, A.; Messner, P.; Sára, M.; Sleytr, U.B.; Beveridge, T.J. S-layered Aneurinibacillus and Bacillus spp. are susceptible to the lytic action of Pseudomonas aeruginosa membrane vesicles. J. Bacteriol. 1998, 180, 2306–2311. [PubMed] 13. Evans, A.G.L.; Davey, H.M.; Cookson, A.; Currinn, H.; Cooke-Fox, G.; Stanczyk, P.J.; Whitworth, D.E. Predatory activity of Myxococcus xanthus outer-membrane vesicles and properties of their hydrolase cargo. Microbiology 2012, 158, 2742–2752. [CrossRef][PubMed] 14. Tashiro, Y.; Yawata, Y.; Toyofuku, M.; Uchiyama, H.; Nomura, N. Interspecies Interaction between Pseudomonas aeruginosa and Other Microorganisms. Microbe. Environ. 2013, 28, 13–24. [CrossRef][PubMed] 15. Schaar, V.; Nordström, T.; Mörgelin, M.; Riesbeck, K. Moraxella catarrhalis outer membrane vesicles carry β-lactamase and promote survival of Streptococcus pneumoniae and Haemophilus influenzae by inactivating amoxicillin. Antimicrob. Agents Chemother. 2011, 55, 3845–3853. [CrossRef] Int. J. Mol. Sci. 2019, 20, 5577 19 of 21

16. Kim, S.W.; Park, S.B.; Im, S.P.; Lee, J.S.; Jung, J.W.; Gong, T.W.; Lazarte, J.M.S.; Kim, J.; Seo, J.-S.; Kim, J.-H.; et al. Outer membrane vesicles from β-lactam-resistant Escherichia coli enable the survival of β-lactam-susceptible E. coli in the presence of β-lactam antibiotics. Sci. Rep. 2018, 8, 5402. [CrossRef] 17. Kulkarni, H.M.; Jagannadham, M.V. Biogenesis and multifaceted roles of outer membrane vesicles from Gram-negative bacteria. Microbiology 2014, 160, 2109–2121. [CrossRef] 18. Joo, H.-S.; Fu, C.-I.; Otto, M. Bacterial strategies of resistance to antimicrobial peptides. Philos. Trans. R. Soc. B Biol. Sci. 2016, 371, 20150292. [CrossRef] 19. Tashiro, Y.; Ichikawa, S.; Shimizu, M.; Toyofuku, M.; Takaya, N.; Nakajima-Kambe, T.; Uchiyama, H.; Nomura, N. Variation of Physiochemical Properties and Cell Association Activity of Membrane Vesicles with Growth Phase in Pseudomonas aeruginosa. Appl. Environ. Microbiol. 2010, 76, 3732–3739. [CrossRef] 20. Hasegawa, Y.; Futamata, H.; Tashiro, Y. Complexities of cell-to-cell communication through membrane vesicles: Implications for selective interaction of membrane vesicles with microbial cells. Front. Microbiol. 2015, 6, 633. [CrossRef] 21. Augustyniak, D.; Seredy´nski,R.; McClean, S.; Roszkowiak, J.; Roszniowski, B.; Smith, D.L.; Drulis-Kawa, Z.; Mackiewicz, P. Virulence factors of Moraxella catarrhalis outer membrane vesicles are major targets for cross-reactive antibodies and have adapted during evolution. Sci. Rep. 2018, 8, 4955. [CrossRef][PubMed] 22. Zhai, B.; Zhou, H.; Yang, L.; Zhang, J.; Jung, K.; Giam, C.Z.; Xiang, X.; Lin, X. Polymyxin B, in combination with fluconazole, exerts a potent fungicidal effect. J. Antimicrob. Chemother. 2010, 65, 931–938. [CrossRef] [PubMed] 23. Ha, K.C.; White, T.C. Effects of Azole Antifungal Drugs on the Transition from Yeast Cells to Hyphae in Susceptible and Resistant Isolates of the Pathogenic Yeast Candida albicans. Antimicrob. agents Chemother. 1999, 43, 763–768. [CrossRef][PubMed] 24. Epand, R.M.; Walker, C.; Epand, R.F.; Magarvey, N.A. Molecular mechanisms of membrane targeting antibiotics. Biochim. Biophys. Acta-Biomembr. 2016, 1858, 980–987. [CrossRef][PubMed] 25. Cowen, L.E.; Sanglard, D.; Howard, S.J.; Rogers, P.D.; Perlin, D.S. Mechanisms of antifungal drug resistance. Cold Spring Harb. Perspect. Med. 2015, 5. [CrossRef] 26. Andersson, D.I.; Hughes, D.; Kubicek-Sutherland, J.Z. Mechanisms and consequences of bacterial resistance to antimicrobial peptides. Drug Resist. Updat. 2016, 26, 43–57. [CrossRef] 27. Tashiro, Y.; Hasegawa, Y.; Shintani, M.; Takaki, K.; Ohkuma, M.; Kimbara, K.; Futamata, H. Interaction of Bacterial Membrane Vesicles with Specific Species and Their Potential for Delivery to Target Cells. Front. Microbiol. 2017, 8, 571. [CrossRef] 28. Yousfi, H.; Ranque, S.; Rolain, J.-M.; Bittar, F. In vitro polymyxin activity against clinical multidrug-resistant fungi. Antimicrob. Resist. Infect. Control. 2019, 8, 66. [CrossRef] 29. Scorzoni, L.; de Paula e Silva, A.C.A.; Marcos, C.M.; Assato, P.A.; de Melo, W.C.M.A.; de Oliveira, H.C.; Costa-Orlandi, C.B.; Mendes-Giannini, M.J.S.; Fusco-Almeida, A.M. Antifungal Therapy: New Advances in the Understanding and Treatment of Mycosis. Front. Microbiol. 2017, 08, 36. [CrossRef] 30. Kathiravan, M.K.; Salake, A.B.; Chothe, A.S.; Dudhe, P.B.; Watode, R.P.; Mukta, M.S.; Gadhwe, S. The biology and chemistry of antifungal agents: A review. Bioorg. Med. Chem. 2012, 20, 5678–5698. [CrossRef] 31. Tatsumi, Y.; Nagashima, M.; Shibanushi, T.; Iwata, A.; Kangawa, Y.; Inui, F.; Siu, W.J.J.; Pillai, R.; Nishiyama, Y. Mechanism of action of efinaconazole, a novel triazole antifungal agent. Antimicrob. Agents Chemother. 2013, 57, 2405–2409. [CrossRef][PubMed] 32. Chen, S.C.A.; Lewis, R.E.; Kontoyiannis, D.P. Direct effects of non-antifungal agents used in cancer chemotherapy and organ transplantation on the development and virulence of Candida and Aspergillus species. Virulence 2011, 2, 280–295. [CrossRef][PubMed] 33. Hsu, L.-H.; Wang, H.-F.; Sun, P.-L.; Hu, F.-R.; Chen, Y.-L. The antibiotic polymyxin B exhibits novel antifungal activity against Fusarium species. Int. J. Antimicrob. Agents 2017, 49, 740–748. [CrossRef][PubMed] 34. Villar, C.C.; Kashleva, H.; Dongari-Bagtzoglou, A. Role of Candida albicans polymorphism in interactions with oral epithelial cells. Oral Microbiol. Immunol. 2004, 19, 262–269. [CrossRef][PubMed] 35. Murzyn, A.; Krasowska, A.; Augustyniak, D.; Majkowska-Skrobek, G.; Łukaszewicz, M.; Dziadkowiec, D. The effect of Saccharomyces boulardii on Candida albicans-infected human intestinal cell lines Caco-2 and Intestin 407. FEMS Microbiol. Lett. 2010, 310, 17–23. [CrossRef] 36. Allison, D.L.; Willems, H.M.E.; Jayatilake, J.A.M.S.; Bruno, V.M.; Peters, B.M.; Shirtliff, M.E. Candida–Bacteria Interactions: Their Impact on Human Disease. Microbiol. Spectr. 2016, 4. [CrossRef] Int. J. Mol. Sci. 2019, 20, 5577 20 of 21

37. Harriott, M.M.; Noverr, M.C. Importance of Candida-bacterial polymicrobial biofilms in disease. Trends Microbiol. 2011, 19, 557–563. [CrossRef] 38. Kerr, J.R.; Taylor, G.W.; Rutman, A.; Høiby, N.; Cole, P.J.; Wilson, R. Pseudomonas aeruginosa pyocyanin and 1-hydroxyphenazine inhibit fungal growth. J. Clin. Pathol. 1999, 52, 385–387. [CrossRef] 39. Gaddy, J.A.; Tomaras, A.P.; Actis, L.A. The Acinetobacter baumannii 19606 OmpA protein plays a role in biofilm formation on abiotic surfaces and in the interaction of this pathogen with eukaryotic cells. Infect. Immun. 2009, 77, 3150–3160. [CrossRef] 40. Krüger, W.; Vielreicher, S.; Kapitan, M.; Jacobsen, I.D.; Niemiec, M.J. Fungal-Bacterial Interactions in Health and Disease. Pathogens 2019, 8, 70. [CrossRef] 41. Roux, D.; Gaudry, S.; Khoy-Ear, L.; Aloulou, M.; Phillips-Houlbracq, M.; Bex, J.; Skurnik, D.; Denamur, E.; Monteiro, R.C.; Dreyfuss, D.; et al. Airway Fungal Colonization Compromises the Immune System Allowing Bacterial Pneumonia to Prevail. Crit. Care Med. 2013, 41, e191–e199. [CrossRef][PubMed] 42. Venkatesh, M.P.; Rong, L. Human recombinant lactoferrin acts synergistically with antimicrobials commonly used in neonatal practice against coagulase-negative staphylococci and Candida albicans causing neonatal sepsis. J. Med. Microbiol. 2008, 57, 1113–1121. [CrossRef][PubMed] 43. Silva, P.M.; Gonçalves, S.; Santos, N.C. Defensins: Antifungal lessons from eukaryotes. Front. Microbiol. 2014, 5, 97. [CrossRef][PubMed] 44. Swidergall, M.; Ernst, J.F. Interplay between Candida albicans and the antimicrobial peptide armory. Eukaryot. Cell 2014, 13, 950–957. [CrossRef][PubMed] 45. Li, J.; Koh, J.-J.; Liu, S.; Lakshminarayanan, R.; Verma, C.S.; Beuerman, R.W. Membrane Active Antimicrobial Peptides: Translating Mechanistic Insights to Design. Front. Neurosci. 2017, 11, 73. [CrossRef][PubMed] 46. Goh, Y.H.; Goode, R.L. State of the Art Review: Current Status of Topical Nasal Antimicrobial Agents. Laryngoscope 2000, 110, 875–880. [CrossRef] 47. Ragioto, D.A.M.T.; Carrasco, L.D.M.; Carmona-Ribeiro, A.M. Novel gramicidin formulations in cationic lipid as broad-spectrum microbicidal agents. Int. J. Nanomed. 2014, 9, 3183–3192. 48. Tan, T.T.; Mörgelin, M.; Forsgren, A.; Riesbeck, K. Haemophilus influenzae Survival during Complement-Mediated Attacks Is Promoted by Moraxella catarrhalis Outer Membrane Vesicles. J. Infect. Dis. 2007, 195, 1661–1670. [CrossRef] 49. Baumgarten, T.; Sperling, S.; Seifert, J.; von Bergen, M.; Steiniger, F.; Wick, L.Y.; Heipieper, H.J. Membrane vesicle formation as a multiple-stress response mechanism enhances Pseudomonas putida DOT-T1E cell surface hydrophobicity and biofilm formation. Appl. Environ. Microbiol. 2012, 78, 6217–6224. [CrossRef] 50. MacDonald, I.A.; Kuehna, M.J. Stress-induced outer membrane vesicle production by Pseudomonas aeruginosa. J. Bacteriol. 2013, 195, 2971–2981. [CrossRef] 51. Yun, S.H.; Park, E.C.; Lee, S.Y.; Lee, H.; Choi, C.W.; Yi, Y.S.; Ro, H.J.; Lee, J.C.; Jun, S.; Kim, H.Y.; et al. Antibiotic treatment modulates protein components of cytotoxic outer membrane vesicles of multidrug-resistant clinical strain, Acinetobacter baumannii DU202. Clin. Proteom. 2018, 15, 28. [CrossRef][PubMed] 52. Park, K.S.; Choi, K.H.; Kim, Y.S.; Hong, B.S.; Kim, O.Y.; Kim, J.H.; Yoon, C.M.; Koh, G.Y.; Kim, Y.K.; Gho, Y.S. Outer membrane vesicles derived from Escherichia coli induce systemic inflammatory response syndrome. PLoS ONE 2010, 5, e11334. [CrossRef][PubMed] 53. Shah, B.; Sullivan, C.J.; Lonergan, N.E.; Stanley, S.; Soult, M.C.; Britt, L.D. Circulating bacterial membrane vesicles cause sepsis in rats. Shock 2012, 37, 621–628. [CrossRef][PubMed] 54. Park, K.S.; Lee, J.; Jang, S.C.; Kim, S.R.; Jang, M.H.; Lötvall, J.; Kim, Y.K.; Gho, Y.S. Pulmonary inflammation induced by bacteria-free outer membrane vesicles from Pseudomonas aeruginosa. Am. J. Respir. Cell Mol. Biol. 2013, 49, 637–645. [CrossRef][PubMed] 55. Namork, E.; Brandtzaeg, P. Fatal meningococcal septicaemia with “blebbing” meningococcus. Lancet 2002, 360, 1741. [CrossRef] 56. Wayne, P.A. Reference method for broth dilution antifungal susceptibility testing of yeasts, Approved standard-second edition. Clsi Doc. M27-A2 2002. 57. Augustyniak, D.; Jankowski, A.; Mackiewicz, P.; Skowyra, A.; Gutowicz, J.; Drulis-Kawa, Z. Innate immune properties of selected human neuropeptides against Moraxella catarrhalis and nontypeable Haemophilus influenzae. BMC Immunol. 2012, 13, 24. [CrossRef] Int. J. Mol. Sci. 2019, 20, 5577 21 of 21

58. Arabski, M.; Lisowska, H.; Lankoff, A.; Davydova, V.N.; Drulis-Kawa, Z.; Augustyniak, D.; Yermak, I.M.; Molinaro, A.; Kaca, W. The properties of chitosan complexes with smooth and rough forms of lipopolysaccharides on CHO-K1 cells. Carbohydr. Polym. 2013, 97, 284–292. [CrossRef] 59. Orwa, J.A.; Van Gerven, A.; Roets, E.; Hoogmartens, J. Liquid chromatography of polymyxin B sulphate. J. Chromatogr. A 2000, 870, 237–243. [CrossRef] 60. Inka, B. Glycobiology Protocols; Humana Press: Totowa, NJ, USA, 2006. 61. Tsai, C.-M.; Frasch, C.E. A sensitive silver stain for detecting lipopolysaccharides in polyacrylamide gels. Anal. Biochem. 1982, 119, 115–119. [CrossRef]

© 2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).